A Hybrid Intelligent Framework for Wind Power Forecasting Engine

Following the growing wind energy integration into the grid, several wind power forecasting technique have been reported in the literature in recent years. This paper presents an advanced hybrid wind power forecasting methodology based on the combination of three different techniques: signal processing, artificial intelligence, and data mining. The signal processing component primarily filters out stochastic wind power time series data. An Adaptive Neuro Fuzzy Inference system (ANFIS), a neuro-fuzzy tool, is applied as a forecasting engine. A data mining tool, support vector machine (SVM) classifier is used to reduce the wind power forecasting error. Finally, grid search (GS) algorithm is applied to optimize the SVM parameters for improving the wind power forecasting performance. The key feature of this paper is to find out a simplified way to forecast wind power without taking into account various input parameters. The proposed wind power forecasting strategy is applied to real-life data from wind power producers in Alberta, Canada. The presented numerical results demonstrate the efficiency of the proposed strategy compared to some other existing wind power forecasting methods.